own right. Why then should it need revision? Why tamper with a work that has been enjoyed and admired, not to say revered, by several generations of readers throughout the English-speaking world?
The answer is that the original French edition from which Scott Moncrieff worked (the “abominable” edition of the
Nouvelle Revue Française
, as Samuel Beckett described it in a marvellous short study of Proust which he published in 1931) was notoriously imperfect. This was not so much the fault of the publishers and printers as of Proust’s methods of composition. Only the first volume (
Du côté de chez Swann
) of the novel as originally conceived—and indeed written—was published before the 1914–1918 war. The second volume was set up in type, but publication was delayed, and moreover by that time Proust had already begun to reconsider the scale of the novel; the remaining eight years of his life (1914–1922) were spent in expanding it from its original 500,000 words to more than a million and a quarter. The margins of proofs and typescripts were covered with scribbled corrections and insertions, often overflowing on to additional sheets which were glued to the galleys or to one another to form interminable strips—what Françoise in the novel calls the narrator’s
“paperoles.”
The unravelling and deciphering of these copious additions cannot have been an enviable task for editors and printers.
Furthermore, the last three sections of the novel (
La prisonnière, La fugitive
—originally called
Albertine disparue
—and
Le temps retrouvé
) had not yet been published at the time of Proust’s death in November 1922 (he was still correcting a typed copy of
La prisonnière
on his deathbed). Here the original editors had to take it upon themselves to prepare a coherent text from a manuscript littered with sometimes hasty corrections, revisions and afterthoughts and leaving a number of unresolved contradictions, obscurities and chronological inconsistencies. As a result of all this the original editions—even of the volumes published in Proust’s lifetime—pullulate with errors, misreadings and omissions.
In 1954 a revised three-volume edition of
À la recherche
was published in Gallimard’s Bibliothèque de la Pléiade. The editors, M. Pierre Clarac and M. André Ferré, had been charged by Proust’s heirs with the task of “establishing a text of his novel as faithful as possible to his intentions.” With infinite care and patience they examined all the relevant material—manuscripts, notebooks, typescripts, proofs, as well as the original edition—and produced what is generally agreed to be a virtually impeccable transcription of Proust’s text. They scrupulously avoided the arbitrary emendations, the touchings-up, the wholesale reshufflings of paragraphs in which the original editors indulged, confining themselves to clarifying the text wherever necessary, correcting errors due to haste or inadvertence, eliminating careless repetitions and rationalising the punctuation (an area where Proust was notoriously casual). They justify and explain their editorial decisions in detailed critical notes, occupying some 200 pages over the three volumes, and print all the significant variants as well as a number of passages that Proust did not have time to work into his book.
The Pléiade text differs from that of the original edition, mostly in minor though none the less significant ways, throughout the novel. In the last three sections (the third Pléiade volume) the differences are sometimes considerable. In particular, MM. Clarac and Ferré have included a number of passages, sometimes of a paragraph or two, sometimes of several pages, which the original editors omitted for no good reason.
The present translation is a reworking, on the basis of the Pléiade edition, of Scott Moncrieff’s version of the first six sections of
À la recherche
—or the first eleven volumes of the twelve-volume English edition. A